Published: February 7, 2017 |
Source:
EE Times |
Author:
Mark Campbell
AMD’s Ryzen CPUs use 10% less die area than their Intel equivalents

AMD’s Ryzen CPUs use 10% less die area than their Intel equivalents
Â
AMD’s Ryzen CPUs use 10% less die area than their Intel equivalents, potentially allowing AMD to produce their Ryzen CPUs more cheaply than Intel.Â
Â
This news clearly shows that AMD has a highly competitive CPU design that can fit into a very small area, despite using Globafoundries’ technically inferior 14nm FinFET process. On thing to note is that Intel’s Kaby Lake CPUs are using a refined version of Intel’s 14nm process and their Skylake CPU design. At this time it is unknown how factors other than die-size will affect AMD’s Ryzen CPUs and their production costs, as this is where yields and Glofabfoundries’ profit margins will come into play. Â
Â
Looking at these basic core specifications AMD’s Ryzen chip could be seen as slightly superior, with 512KB of L2 cache rather than the 256KB per core that Kaby Lake features. That being said there is a lot more to CPU performance than cache sizes and core counts, making it impossible to tell how well AMD’s Ryzen CPUs will perform compared to Kaby Lake. Â Â
Â
 Â
 Â
Â
It is clear from this data that AMD’s Ryzen CPUs are very well designed and make very good use of die space, though it remains to be seen how all this will affect the CPU’s competitiveness in the market. Â
In EE Times’ recent article on this topic, it was stated that “Analysts and even Intel engineers in the session said the Zen core is clearly competitive”, which is a highly impressive feat for AMD. Let’s hope that AMD’s Ryzen CPUs will continue to impress when they hit the market next month.Â
Â
You can join the discussion on AMD’s Ryzen CPUs on the OC3D Forums.Â
Â